Here is a new version using H2 Generator.
The level is 1mW into 32 Ohm = +0dB.
R1 should be adjusted to give like 7-8 Volt across the JFET.
As can be seen we now have lots of H2 at this lower level.
The level is 1mW into 32 Ohm = +0dB.
R1 should be adjusted to give like 7-8 Volt across the JFET.
As can be seen we now have lots of H2 at this lower level.
Actually, two OPA1655 in parallel drive 32 Ω pretty easily. That's what I do in the HP-2. You can see the performance with 32 Ω load is the Specifications and Performance Graphs tabs of the product page: https://neurochrome.com/products/hp-2
The OPA593 provides 250 mA output current, so that might be an option for a headphone driver as well. It's pretty noisy, though. YMMV.
Tom
Since my Grados produce 100 dB SPL with 1 mW, they will produce 110 dB SPL with 10 mW. A factor of 10 in power is 10 dB. A factor of 10 in voltage is a factor of 100 in power so is 20 dB.
I see you like to find things to make fun of, but when did I ever say anything about using a standard opamp? While most opamps would be terrible at it, there are some opamps that are happy with 32 Ohm loads and can drive them with low distortion. The njm4556 is a low-cost and common example. I am unfamiliar with the opa1655, which is why one of the first things I posted in this thread was my curiosity about how well it would handle a 32-Ohm load.
I usually use discrete myself - just because I find it to be more fun for some reason. I have a few njm4556 opamps laying around in case I want a quick-and-dirty $1 driver, but have yet to use one except for experiments on a breadboard.
NJM4556 rated at 70ma output current
OPA1655 rated at 100ma output current
There is other variables to how well opamp can drive low impedance
with low distortion. But output current be a big factor.
1655 is rail to rail, good slew rate , bandwidth and low distortion.
Cell phones have headphone drivers with less and run on 3.3 / 5 volts
So a opamp with 10 , 15 or 30 volts. Has much more potential.
TPA16111A2 and other countless examples of stereo or 2 channels
for single supply. Usually have thermal shutdown and other features
to make them reliable.
Way less than a 1655 and a LM7171 would smoke them all
OPA1655 rated at 100ma output current
There is other variables to how well opamp can drive low impedance
with low distortion. But output current be a big factor.
1655 is rail to rail, good slew rate , bandwidth and low distortion.
Cell phones have headphone drivers with less and run on 3.3 / 5 volts
So a opamp with 10 , 15 or 30 volts. Has much more potential.
TPA16111A2 and other countless examples of stereo or 2 channels
for single supply. Usually have thermal shutdown and other features
to make them reliable.
Way less than a 1655 and a LM7171 would smoke them all
Heh. No kidding! I tried making a programmable current source with that for testing high-voltage regulators. I turned quite a few parts into slag that way. The fireworks were spectacular. 🙂
I'm sure you could make a headphone amp with an LM7171 (or its cousin, the LM6171) and get it to be stable at one specific load impedance. But neither of those would be my first choice in a circuit where the (reactive) load is determined by the end user.
There are quite a few buffers and headphone amp ICs that will happily turn into oscillators if you let them. I had an OPA1611/LME49600 combo oscillating in the 100-200 MHz region during the development of the HP-1.
I'm guessing you mean a TPA6120A2. That's a pretty solid chip. I've played with it in a headphone amp configuration and it performs very well. I haven't needed any sort of thermal shutdown even with a 12 Ω load. I have about 60 of such amps out in the field and have yet to have a warranty return.
As you probably know, it's half of an OPA1656. Quite a few use the OPA1656. It's a very nice opamp.
Tom
Composite has its challenges, but rewarding when stable.
7171 performance rather amazing.
But a can be tricky one.
not 15 or 60 MHz it is 200 MHz so
decoupling caps are not 100n and 10u
Decoupling caps are 10n and 2.2u
Most are use to getting away with certain
circuit designs because amps are usually compensated.
7171 is uncompensated
and stable at +2 or -1 gain
So assuming, massive DC offset, decoupling issues
and using at unity gain. The outcome you got would
be likely.
Not picking on you though
Same thing would have happened to me actually. LOL
I played with it in Sim first and realized the differences.
Poked around the datasheet and actually enjoyed going
back to basics and playing with circuits for uncompensated
amps. More mentioned it, because the performance is unreal
and would rather see people get benefit from it, not frustration